Plans for Merging X3D AR Proposals
This page is for discussing plans for merging X3D AR proposals, compared in Comparison of X3D AR Proposals.
These are the steps we will take as a process of merging the X3D AR proposals:
1. Discuss general strategy/policy/guidelines
- General design guidelines in Web3D consortium level? → Not in explicit form, but there are in the concept sections of node specifications.
- Notes from Don and Dick:
- When a new field is added to a node, carefully choose a default value for backward compatibility.
- For consistency, mixing multiple functions into a single node is not recommended.
- Device independence is taken for granted.
- For identifying devices, URNs could be a proper way to describe them.
- Metric for names - We have the right name if nobody asks about it anymore.
- Start with designing abstract functionality first and then move on to the node specification.
- Build examples/use cases and use them as a test for integrity.
- Rule of thumb
- From the scene writer's point of view, preferably less code for commonly/frequently used features, while detail control available for special cases.
- From the user's (viewer's) point of view, the scene should be adopted to the hardware/software environment (tracker, camera device, browser, etc.) s/he has. In other words, scene writer should not specify hardware/software environment in the scene, which is on the users' side.
2. Produce a merged proposal for each functional components
- Investigate each functional features stepwise:
- Camera video stream image into the scene (texture and background)
- Tracking (including support for general tracking devices)
- Camera calibration (viewpoints)
- Others (color-keying, depth occlusion)
3. Check Integrity of the merged proposal
- Check and resolve conflicts between individual functional components
- Merge overlapping features between individual functional components
4. Specification writing